Faster Was Never the Point

In March 2026, Lloyd’s of London made a bold decision: to transition away from Blueprint Two, its six-year plan to build one central digital system for the entire insurance market. Lloyd’s leadership shared what many people in the market had quietly felt for a while, one giant system was never going to make every company smart on its own.

For companies that had been waiting for this platform before investing in their own technology, the news could have felt like a setback. But for the companies that hadn’t waited, it barely mattered. They had already stopped waiting for someone else to bring them intelligence. They were building it themselves, one small improvement at a time.

The real opportunity in insurance today isn’t about technology. It’s about knowing the difference between doing things faster and doing things smarter.

Automation Got Us This Far

For many years, technology in insurance has mostly meant automation computers doing the same repetitive steps a person used to do, just faster and without breaks. This has helped a great deal. Take a Bordereau a routine batch of policy and claims data insurers exchange with their partners every month. In one recent case, automating this process cut the handling time for each file from around 30–35 minutes down to under five, roughly an 85–90% improvement in speed. Across the wider industry, the share of claims settled without a person touching them at all has climbed from around 10–15% five years ago to as high as 70–90% today in the most AI-enabled operations. None of this progress would have been possible without automation, and the industry should be proud of how far it has come.

But automation has a limit, and many companies are reaching it. A 2025 survey of London market firms found that almost three-quarters now use AI to pull data out of documents, but only 14% have gone further and used it to actually support an underwriting decision. Most insurers can already pull information out of forms and documents automatically. Far fewer are using that information to help make real decisions, judging whether something looks risky, spotting a mistake before it becomes a problem, or explaining clearly why a number doesn’t look right. That gap, between collecting information and truly using it well, is where the next big opportunity lies.

What Intelligence Really Means

Automation follows a rule. Intelligence understands when a rule doesn’t quite fit. A computer can notice that a number looks unusual. It takes something closer to real understanding to explain why that number is unusual, clearly enough that someone can act on it right away instead of spending an hour figuring it out. That difference, between noticing something and understanding it, is what separates automation from intelligence.

This shift is already happening across insurance. Some tools now check that information is consistent and accurate before small errors become bigger problems. Others can review straightforward cases and make simple decisions on their own, so people only need to step in for the more complex ones. None of this replaces people. It changes what people get to spend their time on.

Helping People, Not Replacing Them

This matters because insurance is facing a real challenge: roughly 400,000 experienced insurance professionals in the US alone are expected to retire by the end of 2026, according to US government workforce projections⁴ — and years of knowledge are not easy to replace overnight. Picture an underwriter who has spent twenty years in the market. They can sense that a risk doesn’t feel right, even when every number on the page checks out. That kind of judgment can’t be handed over on day one — but a good intelligence tool can help a newer colleague spot the same warning signs faster, building that instinct in years rather than decades. Automation alone cannot fix this. The companies doing well in 2026 are not the ones spending the most money on technology. They are the ones using it to help their people do their best work — freeing up time once spent on repetitive checks so people can focus on the decisions only they can make, and helping newer team members learn faster by giving them the same context an experienced colleague once carried in their head.

The question has changed. Most companies already answered “what can we automate?” years ago. The better question now is: where can a smart system support our people and help them move faster, while people still make the final call? That is a more positive and more human way to think about technology, and it is where the real value is being created.

The Lesson from Lloyd’s

The lesson from Lloyd’s isn’t that one big idea failed. It’s that real progress doesn’t need to wait for one perfect plan. It happens step by step, inside the companies willing to see technology as a helpful partner rather than just a faster way to do the same old work. Lloyd’s has said it will now support the market through smaller, steady improvements instead of one giant system, which really means the pace of progress is now in the hands of individual companies.

The next chapter for insurance won’t be announced by one organisation or one big plan. It is already happening quietly, inside the teams who decided not to wait for someone else to make them smarter.